Create alerts when you need Yarken to monitor a metric and notify the right team when a condition is met.
This page explains how to create one alert from the Alerting page. For editing, activation, manual runs, run history, import, and export, see Manage, Import, and Export Alerts.
Before you start
Confirm that a notification channel has already been configured and enabled.
Alerts need a notification channel so Yarken knows where to send the message when the alert triggers. Depending on the environment setup, channels may include Microsoft Teams, Slack, or other supported destinations.
You should also know:
-
Which metric the alert should monitor
-
Whether the alert should check for a threshold or missing data
-
Whether the alert should run on a schedule or from an application event
-
Which channel should receive the message
-
What action the recipient should take when the alert triggers
Create an alert
To create an alert:
-
Navigate to Admin > Alerting
-
Click Add New
-
Complete the New Alert form
-
Save the alert
-
Run the alert manually to test it where needed
-
Activate the alert when it is ready for ongoing monitoring
The form is organized into four main sections: alert details, metric selection, trigger configuration, and notification setup.
Add alert details
Use the first section to define the alert name and purpose.
|
Field |
What to enter |
|---|---|
|
Alert name |
A clear name that describes the metric, scope, and trigger |
|
Description |
A short explanation of what the alert checks and why it matters |
Use names that are easy to understand in the Alerting list.
Examples:
-
AWS Monthly Spend
-
Check Microsoft Spend
-
Check monthly load
-
Check M-FILES spend
Avoid vague names such as Spend alert or Monthly check.
Select a metric
The metric defines what the alert monitors.
To select a metric:
-
Open the Metric dropdown
-
Select the metric the alert should monitor
-
Review any additional metric options that appear
Metric availability depends on the data and configuration in the environment.
Common alert metrics may include:
-
Monthly spend
-
Application spend
-
Vendor spend
-
Cost pool spend
-
Tower spend
-
Cloud spend
-
Variance percentage
-
Data load status
Choose the metric that maps directly to the condition you want to monitor.
Configure the trigger condition
The trigger condition defines when the alert should notify the selected channel.
Select the Condition type, then complete the fields that appear for that condition.
Common condition types include:
|
Condition type |
Use it when |
|---|---|
|
Threshold |
The alert should trigger when a value crosses a defined amount or percentage |
|
Absence of value |
The alert should trigger when expected data is missing |
Set up a threshold alert
Use a threshold alert when a metric should trigger a notification based on value.
To configure a threshold:
-
Select Threshold as the condition type
-
Select the threshold position, such as above, below, within, or outside a range
-
Enter the threshold value
-
Confirm the metric period or scope if the form shows additional options
Threshold alerts are useful for:
-
Spend exceeding a defined limit
-
Vendor spend passing a review threshold
-
Application spend moving outside an expected range
-
Variance exceeding a tolerance level
-
Cloud spend crossing a monthly control point
Set up an absence alert
Use an absence alert when expected data should exist but does not.
To configure an absence alert:
-
Select Absence of value as the condition type
-
Select the metric or data check to monitor
-
Confirm the period, scope, or expected value rules if shown
Absence alerts are useful for:
-
Monthly spend not being loaded by a required date
-
Expected data missing from a reporting period
-
Upload or ingestion checks
-
Month-end readiness checks
Configure the trigger schedule
The trigger schedule defines when the alert runs.
Select the Trigger type, then complete the fields that appear.
|
Trigger type |
Use it when |
|---|---|
|
Schedule |
The alert should run at a defined cadence |
|
Application event |
The alert should run when a system event occurs |
Use a scheduled trigger
Use a scheduled trigger for recurring checks.
To configure a scheduled trigger:
-
Select Schedule as the trigger type
-
Choose the frequency or schedule options shown in the form
-
Align the schedule to the reporting cycle or review cadence
Scheduled alerts work well for daily, weekly, monthly, or month-end checks.
Use an application event trigger
Use an application event trigger when the alert should run after specific system activity.
To configure an event trigger:
-
Select Application event as the trigger type
-
Select the event that should run the alert
-
Confirm the alert condition still matches the event timing
Event-based alerts work well for checks tied to uploads, data loads, or application processing activity.
Configure notification
The notification section controls where the alert is sent and what message is delivered.
To configure notification:
-
Select the Notification channel
-
Enter the Channel message
-
Confirm the message gives the recipient enough context to act
A useful channel message should include:
-
The metric being monitored
-
The condition that triggered the alert
-
The affected period or scope
-
The expected action or review owner
Example:
AWS monthly spend has exceeded the configured threshold for the current month. Review provider spend and confirm whether follow-up is required.
Save, test, and activate
After completing the form, click Save.
After saving, use the Alerting list to:
-
Run the alert manually
-
Review the run result
-
Check run history if needed
-
Activate the alert when it is ready to run automatically
Testing helps confirm the metric, condition, trigger, notification channel, and message are configured correctly.
Common mistakes to avoid
Avoid these issues when creating alerts:
-
Creating an alert before the notification channel is configured
-
Using vague alert names that are hard to recognize later
-
Setting thresholds without confirming the metric unit or period
-
Sending messages that do not explain what action is required
-
Activating an alert before testing it
-
Using a scheduled alert when the check depends on a data-load event
Clear configuration reduces noise and improves response quality.
Troubleshooting alert creation
If the alert cannot be saved or does not behave as expected, check:
-
Required fields are completed
-
A valid metric is selected
-
The condition type is configured correctly
-
Threshold values are entered in the expected format
-
The trigger type is selected
-
The schedule or event is valid
-
The notification channel is enabled
-
The channel message is populated
-
The alert has been activated if it should run automatically
If the alert runs but no notification is received, review the notification channel configuration first.
Next step
Related content